How to solve physics problems

Unlike history or language which has more to do with remembering facts (at least at first), learning physics is more like learning to DO something (play a piano, draw a picture, program a computer for example). You don't have to remember very much (fortunate for me since I have a terrible memory) but you have to know how to use it.

Students often look at a problem, they say 'gosh I don't know the answer to that' and they freeze. But no one (not even an instructor) can look at a problem and know the answer. You have to work through from what you do know to FIND the answer. This bothers students, they feel lost because they don't see how they are going to get to the answer when they start. But generally if you just muddle along, applying equations and concepts that you are sure are right, eventually some route to the answer will pop out in front of you.

The following recipe may help you get started:

Draw a picture. This may clarify what is going on and gives you something to do while your brain gets past the panic.
Write down a list of things you know about the problem (is velocity given? acceleration? any hidden information such as 'the object starts from rest' which is equivalent to saying the initial velocity is zero?)
Write down a list of potential equations that might apply to the problem, keeping in mind what the terms mean and what the equation 'says' in English. (For example x=vt is only good for average velocity; if there is acceleration then x=vt+1/2at² which means that you go a distance x=vt (where v is the initial velocity) plus a bit more (+1/2at²) because you are speeding up with acceleration, a)
Divide and conquer. Can you divide the problem into stages? Suppose a car has acceleration for 4s and then coasts. Probably you can treat this as two separate problems, one with acceleration, the other without. If you need to find the total distance, find the distance for each part and then add. The great thing about those equations is they are reusable, you can apply them over and over, even on the same problem.
Don't randomly substitute numbers into equations, hoping you get a useful answer. Stick with equations and numbers that make sense. It is important to know what the equations mean, not just be able to recall what they look like. And also, equations are mathematical sentences that embody a physical concept or principle. The concept is actually more important than the equation but when it comes down to finding an answer the equation is useful. But knowing the concept behind the equation makes it easier to select the right equation.
Keep track of units (acceleration is m/s² not m/s). This may help you discover errors in your math. It may also keep you from giving an answer that doesn't make sense (if the problem asks for speed and you come up with a number in seconds, you aren't done yet).
When you think you are done, think about whether the answer makes sense. Does a speed of 1000m/s for a car make sense? Only if you are launching it into space!
Often there is more than one way to get to the right answer. If you can get the same answer two different ways, probably you are on the right track.
Practice. Like learning to ride a bike, you are probably going to fall off a few times. I still occasionally start working a problem in class and get stuck. Most texts have example problems -don't just read them, cover the steps with a piece of paper and try to come up with the steps and then check to see if you are right. Some books have a student study guide with examples; do the same (you can find a Schaum's outline with examples too). Most texts give answers to even (or odd) problems so you can check your work. The instructor probably does some examples in class- don't memorize them, try to understand what the steps are that he left out.
Get a study group together. Three heads are better than one (four or five is probably too many to be efficient). Solve problems together in the study group, talk about each step along the way, giving reasons for each step.

Good Luck!